Gate News message, April 27 — Alibaba’s PAI team has released and open-sourced AgenticQwen, a small-scale agentic language model designed for industrial-grade tool-calling applications. The model comes in two versions: 8B and 30B-A3B. Trained through an innovative “dual data flywheel” reinforcement learning framework, AgenticQwen achieves near-trillion-parameter model-level agentic capabilities while significantly reducing inference costs.
The dual data flywheel mechanism addresses the homogenization problem in traditional synthetic data. The reasoning flywheel automatically generates harder variants from model errors, while the agentic flywheel expands simple linear workflows (such as single booking processes) into multi-branch behavior trees with constraints, rejections, and adversarial conditions, simulating real-world complex decision scenarios. Benchmarks show AgenticQwen-8B scored 47.4 on real tool environment benchmarks (TAU-2 and BFCL-V4), far exceeding the base Qwen3-8B (23.8) and approaching Qwen3-235B (52.0). AgenticQwen-30B-A3B (with only 3B parameters activated) achieved 50.2.
The model has been deployed in internal production systems similar to Manus, significantly narrowing the gap with 235B models in end-to-end inference time. However, the model is limited by a native context length of 40K tokens, which constrains its performance on deep search tasks.
Disclaimer: The information on this page may come from third parties and does not represent the views or opinions of Gate. The content displayed on this page is for reference only and does not constitute any financial, investment, or legal advice. Gate does not guarantee the accuracy or completeness of the information and shall not be liable for any losses arising from the use of this information. Virtual asset investments carry high risks and are subject to significant price volatility. You may lose all of your invested principal. Please fully understand the relevant risks and make prudent decisions based on your own financial situation and risk tolerance. For details, please refer to
Disclaimer.
Related Articles
AI Agent Manfred Forms Company, Gets Ready to Trade Crypto by End of May
AI agent Manfred has formed its own company and already possesses a crypto wallet and business credentials to hire staff, make payments, and conduct operations. The agent is scheduled to begin trading crypto by the end of
GateNews55m ago
AI Agent Manfred Forms Company, Gets Crypto Wallet and Hiring Credentials Ahead of End-May Trading Launch
AI agent Manfred has formed its own company and obtained a crypto wallet along with credentials to hire staff, make payments, and conduct business, according to reports. The agent is scheduled to begin trading crypto by the end of
GateNews6h ago
MoonPay Launches MoonAgents Card, Virtual Mastercard for AI Agents, on Friday
According to The Block, MoonPay launched MoonAgents Card, a virtual Mastercard debit card, on Friday. The card is designed for both AI agents and users, converting stablecoins into fiat at the point of payment and enabling spending at any online merchant globally that accepts Mastercard.
The card i
GateNews12h ago
AI Agent Manfred Forms Company, Prepares to Trade Crypto by End of May
AI agent Manfred has already established a company and obtained a crypto wallet along with credentials to hire staff, make payments, and conduct business. The agent is scheduled to begin crypto trading by the end of
GateNews12h ago
ChatGPT Users Can Now Access Subscriptions on OpenClaw Platform, Sam Altman Announces
According to Sam Altman's announcement today, ChatGPT users can now log in to the OpenClaw agent platform using their ChatGPT accounts and directly access their existing ChatGPT subscriptions on the platform.
GateNews18h ago
AI Agent Manfred Forms Company, Plans to Start Trading Crypto by End of May
AI agent Manfred has formed its own company and obtained a crypto wallet along with credentials to hire staff, make payments, and conduct business, according to reports. The agent plans to begin trading crypto by the end of
GateNews19h ago
